The experiment analysed nearly 24,000 grant applications over five years at the Canadian Institute of Health Research -- Canada's main public medical research funder.

In 2014 the body changed its application process, splitting funding reviews into two separate schemes -- one with an explicit focus on the applicant, the other evaluating the science.

In doing so, they created a "unique natural experiment", according to the authors of the study.

When assessments were based solely on the quality of the science, the gender gap between grants accepted was a mere 0.9 percentage points.

But when the assessments were based on an evaluation of the principal scientists pitching the project, the gap between male and female acceptances grew to 4 percent.

"This shows us that men and women proposed science is evaluated to be of similar quality, but men and women are not evaluated similarly as scientists," said Holly Witteman, associate professor at the Department of Family and Emergency Medicine at Laval University, Quebec.

Witteman said there may be a number of reasons behind this, including individual or systemic biases.

Whether consciously or unconsciously, reviewers may "tend to think as men being better scientists than women," she told AFP.